We comment on the argument by Welzel, Brunkert, Kruse and Inglehart (2021) that theoretically expected associations in nomological networks should take priority over invariance tests in cross-national research. We agree that narrow application of individual tools, such as multi-group confirmatory factor analysis with data that violates the assumptions of these techniques, can be misleading. However, findings that fit expectations of nomological networks may not be free of bias. We present supporting evidence of systematic bias affecting nomological network relationships from a) previous research on intelligence and response styles, b) two simulation studies, and c) data on the choice index from the World Value Survey (WVS). Our main point is that nomological network analysis by itself is insufficient to rule out systematic bias in data. We point out how a thoughtful exploration of sources of biases in cross-national data can contribute to stronger theory development.

我们评论了 Welzel、Brunkert、Kruse 和 Inglehart (2021) 的论点，即理论上预期的规律网络中的关联应该优先于跨国研究中的不变性检验。我们同意个别工具的狭隘应用，例如使用违反这些技术假设的数据的多组验证性因素分析，可能会产生误导。然而，符合规律网络预期的发现可能并非没有偏见。我们从 a) 先前对智力和反应方式的研究，b) 两项模拟研究，以及 c) 来自世界价值调查 (WVS) 的选择指数数据，提供了系统偏差影响法理网络关系的支持证据。我们的主要观点是，法理网络分析本身不足以排除数据中的系统偏差。